Abstract

In 'Reference and Contingency', Monist 62 pp. 161-89, Gareth Evans tried to explain how a priori knowledge of contingent truths is possible by arguing that indexical features of the truths in question make them contingent only in a superficial sense. In 'The Contingent A Priori: Has it Anything to do with Indexicals?', ANALYSIS 46.3, June 1986, pp. 113-7, I suggested that his explanation is inadequate, since a priori knowledge is also possible of deeply contingent truths with no relevant indexical features. Graham Oppy disagrees; in 'Williamson and the Contingent A Priori', ANALYSIS 47.4, October 1987, pp. 188-93, he claims to have detected hidden indexicality in my examples. I argued that beliefs formed by the following method constitute a priori knowledge: Given a valid deduction from the premiss that someone believes that P to the conclusion that P, believe that P. If one puts 'There is at least one believer' for 'P', the a priori knowledge is of a non-indexical contingent truth. It will not be necessary to rehearse the details of this argument in order to explain why I do not find Oppy's objections persuasive